SUMMARY OF WORK In cardiac myocytes, Ca2+ release from RyR in the sarcoplasmic reticulum (SR) is activated by the Ca2+-induced-Ca2+ release (CICR) mechanism. CICR, with its inherent positive feedback, is expected to operate in an "all-or-none" fashion. In order to generate Ca2+ transients of graded amplitude and robust stability, a regulatory mechanism must exist to counteract the regenerative CICR. Several mechanisms, including inactivation, adaptation, and stochastic closing of RyRs have been proposed, but no conclusive evidence has yet been documented. Our recent study has shown that FK506-binding protein (FKBP), an immunophilin and accessory protein of RyR, constitutes a prominent regulator of CICR via shortening the duration of the elementary release events (Ca2+ sparks) and accelerating the desensitization of RyR to Ca2+. However, the primary termination mechanism of CICR remained elusive. In the present study, we probed the termination process of Ca2+ release triggered by L-type Ca2+ channel using a novel fluorescent technique. By combination of a fast, linear Ca2+ indicator, Oregon Green BAPTA 5N, and a high concentration of Ca2+ chelator, EGTA, Ca2+ release was visualized as discrete "Ca2+ spikes" restricted toT tubule-SR junctions, each consisting of single or a few Ca2+ sparks. At 0 mV, Ca2+ spikes occurred and terminated within 40 ms following the onset of voltage clamp pulses. Increasing the open duration and promoting the reopenings of Ca2+ channels with the Ca2+ channel agonists, FPL64176, did not prolong or trigger secondary Ca2+ spikes, even though 2/3 of the SR Ca2+ remained available for release by caffeine. Latency analysis revealed that Ca2+ spikes coincided with the first openings, but not with the reopenings, of L-type Ca2+ channels. Furthermore, after an initial maximal release (e.g., at 0 mV), even a multi-fold increase in unitary Ca2+ current produced by a hyperpolarization step to -120 mV failed to trigger additional release, indicating an absolute refractoriness of RyRs. When the release was submaximal (e.g., at +30 mV), tail currents upon hyperpolarization did activate additional Ca2+ spikes; confocal images revealed that they originated from a different RyRs, i.e., those unfired during depolarization. These results indicate that Ca2+ release is terminated primarily by a highly localized, use-dependent inactivation of RyRs , but not by stochastic closing and adaptation of RyRs or depletion of SR Ca2+ in intact ventricular myocytes.